Releases: gardener/external-dns-management
Releases · gardener/external-dns-management
v0.8.5
[external-dns-management]
🐛 Bug Fixes
- [OPERATOR] use garden cluster identity and add dnsowner for migration (needs Gardener version >= 'v1.22.2' and < 'v1.23.0') (#176, @MartinWeindel)
v0.8.4
[external-dns-management]
🐛 Bug Fixes
- [USER] Deduplication of targets if mutiple CNAME targets are provided (#171, @MartinWeindel)
- [OPERATOR] ensure stable provider result in method DNSProviders.LookupFor (#172, @MartinWeindel)
🏃 Others
- [OPERATOR] Added
external_dns_management_requests_per_zone
andexternal_dns_management_zone_cache_discardings
metrics (#173, @MartinWeindel)
v0.8.3
[external-dns-management]
🐛 Bug Fixes
- [USER] DNS entries without targets are handled as invalid and can be deleted (#170, @MartinWeindel)
v0.8.2
[external-dns-management]
🐛 Bug Fixes
- [OPERATOR] fix premature deletion of DNSEntry if deletion fails in provider (#165, @MartinWeindel)
🏃 Others
- [OPERATOR] The
revisionHistoryLimit
of the dns-controller-managerDeployment
was increased. (#168, @MartinWeindel) - [OPERATOR] log if zone cache is dropped (#166, @MartinWeindel)
- [OPERATOR] Google CloudDNS: use project id as prefix of zone id (#163, @MartinWeindel)
v0.8.1
[external-dns-management]
🐛 Bug Fixes
- [OPERATOR] infoblox: fix panic on setting value for unexported field certPool (#155, @MartinWeindel)
🏃 Others
- [USER] improve error message for 'no domain matching zones' (#158, @MartinWeindel)
- [OPERATOR] Fix: do not remove entry finalizer if provider backend is temporarily unavailable during reconciliation (#162, @MartinWeindel)
- [OPERATOR] consider provider zone on adding entries to zone (#160, @MartinWeindel)
- [OPERATOR] CRDs are not deployed by the helm chart with default values anymore. (#159, @MartinWeindel)
- [OPERATOR] The dns-controller-manager chart does now define a PriorityClass to prevent preemption. (#157, @ialidzhikov)
- [OPERATOR] reduce logging during provider deletion (#153, @MartinWeindel)
v0.8.0
[external-dns-management]
⚠️ Breaking Changes
- [OPERATOR] There is now a shared DNS controller used by all provider types. (#150, @MartinWeindel)
- The values for the command line argument
--controllers
have changed and may have to be adjusted. - Concretely the values
aws-route53
,alicloud-dns
,azure-dns
,google-clouddns
,cloudflare-dns
, ìnfoblox-dns`, - and
openstack-designate
are not supported anymore and are replaced with the single valuecompound
. - The controller can be restricted to specific provider types with the option
--provider-types
. - Additionally many provider specific configuration names are not existing anymore. These settings have to applied to the
compound
controller instead.
- The values for the command line argument
🏃 Others
- [OPERATOR] Added a new Netlify DNS Provider https://open-api.netlify.com/#tag/dnsZone (#147, @afrittoli)
v0.7.22
[external-dns-management]
🏃 Others
- [USER] The list of canonical hosted zones for creating AWS alias targets has been extended by the regions af-south-1, eu-south-1, and us-gov-east-1. The hosted zone ids for regions cn-north-1 and cn-northwest-1 have been fixed. (#142, @MartinWeindel)
- [OPERATOR] AWS Route53: support for chain of credential providers (#149, @MartinWeindel)
- [OPERATOR] infoblox: set TTL on record creation explicitly (#148, @MartinWeindel)
- [OPERATOR] Environment variables can now be set via values.yaml (#139, @mganter)
- [OPERATOR] additional printer columns for DNSEntries, DNSProviders, DNSOwners (#138, @MartinWeindel)
- [OPERATOR] optionally specify default TTL on provider (#137, @MartinWeindel)
v0.7.21
[external-dns-management]
Improvements
- [OPERATOR] set finalizer immediately when DNSEntry becomes valid (fix for "already busy for owner" problems) (#136, @MartinWeindel)
- [OPERATOR] don't set new DNSEntry to invalid if DNSOwner is pending (#135, @MartinWeindel)
- [OPERATOR] infoblox: allow SSL_VERIFY to be specified in the DNSProvider secret (#132, @MartinWeindel)
- [OPERATOR] restrict caching to target namespace for DNS source controllers to reduce logging (#130, @MartinWeindel)
- [OPERATOR] added lastUpdateTime field to status of DNSProvider and DNSEntry (#129, @MartinWeindel)
v0.7.20
[external-dns-management]
Action Required
- [OPERATOR] If dns-controller-manager is used on clusters not created ("shooted") by Gardener and you are using (#123, @MartinWeindel)
- the source controllers with different default and target clusters, you are now required to provide an id for the default cluster explicitly. No non-unique default value is assumed anymore in this case.
- There are two ways to provide it.
- Use the command line option
--kubeconfig.id
or the corresponding valueconfiguration.kubeconfigId
for the helm chart - Or alternatively you can create a configmap
cluster-identity
on the default cluster. Seeexamples/15-cluster-identity.yaml
for concrete structure.
Improvements
- [OPERATOR] use component prefix for prometheus metrics names (#126, @MartinWeindel)
- [OPERATOR] support migration of
resources.gardener.cloud/owners
with the--migration-ids
option (#123, @MartinWeindel)
v0.7.19
[external-dns-management]
Improvements
- [OPERATOR] alicloud-dns/cloudflare-dns/infoblox providers: fix creation of DNSEntry for wildcard domain for the base domain of a hosted zone (#122, @MartinWeindel)
- [OPERATOR] Ready DNSEntries with provider having problem enter State state instead of Error (#117, @mandelsoft)